[daip] [!6831]: AIPS - problem with u-v data after conversion to FITS

[Eric Greisen]

Eric Greisen updated #6831
--------------------------

              Status: Closed (was: Open)
                 Due: - Cleared - (was: 09 July 2015 04:49 AM)

problem with u-v data after conversion to FITS
----------------------------------------------

           Ticket ID: 6831
                 URL: https://help.nrao.edu/staff/index.php?/Tickets/Ticket/View/6831



Usually there are no records completely flagged after processing
and writing to FITS.  The dropping of fully flagged was done for
inputs from CASA in the early EVLA days.  I have made a change  to
cover for CASA (no NX table) and cases like this:

14616.  July 7, 2015          UVLOD, FITLD             Eric
        These tasks read index tables written by FITTP and FITAB.
        When reading the former it is possible to have UV records
        discarded if they are completely flagged (under control of
        DOKEEP).  Added subroutine NXCHCK to both tasks to check if
        there is an index table and if so does the peak vis record
        number match the peak of the vis file just written.  If not,
        the NX table is recomputed (or created new if no input table
        was provided).  This covers for the lack of NX tables from
        CASA and for odd behaviors when records are omitted.
        Moved nowhere.

Eric




------------------------------------------------------
Staff CP:  https://help.nrao.edu/staff